YouTube Video Transcript

v1.0.0

Fetch, summarize, and save YouTube transcripts with timestamp navigation, chapter detection, and searchable content.

1· 492·2 current·2 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The skill's name and description match its runtime instructions: it uses yt-dlp to extract subtitles/transcripts, detects chapters, preserves timestamps, and caches files under ~/youtube-video-transcript/. Requiring yt-dlp (via brew/pip) is exactly what is needed.
Instruction Scope
Instructions stay within the stated purpose (running yt-dlp, parsing VTT/JSON, saving markdown transcripts). They explicitly require user consent for caching and instruct not to store credentials. Two items to be aware of: (1) the skill includes batch/channel/playlist processing patterns which enable mass transcript extraction — legitimate for research but potentially privacy- or copyright-sensitive; (2) it suggests the agent may "offer to automatically transcribe when you share YouTube links," which is a behavioral preference that must be consented to (it grants the agent broader proactive behavior if enabled).
Install Mechanism
Install spec references standard package sources (Homebrew formula 'yt-dlp' and pip as an alternate) — both are expected and traceable. No arbitrary download URLs or archive extraction are used.
Credentials
No environment variables, no credentials, and no external config paths are requested. The skill mentions cookies for age-restricted videos but explicitly says the user must provide them and that the skill will not ask for or store credentials.
Persistence & Privilege
always:false and no special privileges. The skill writes only to a user-owned folder (~/youtube-video-transcript/) and asks for caching consent before storing transcripts. This is proportional to its stated functionality.
Assessment
This skill appears coherent and implements what it advertises, but consider the following before installing: - Installation: it requires yt-dlp (Homebrew or pip). Approve installing yt-dlp from those package sources if you trust them. - Caching & privacy: transcripts are stored under ~/youtube-video-transcript/ only with your explicit consent; refuse caching if you don't want local copies. The skill promises to show and delete cached files on request — verify that behavior after use. - Cookies & auth: the skill may recommend using your browser cookies for age-restricted videos; never paste passwords or tokens into chat. Provide cookies only if you understand the implications and do not expect the skill to store them. - Batch processing risk: the skill includes batch/channel/playlist workflows (mass extraction). This is legitimate for research but can be privacy- or copyright-sensitive; don't enable bulk processing without confirming you have the right to archive that content. - Proactive behavior: it may offer to automatically transcribe when you share YouTube links — only enable that if you want the agent to act on links without an explicit per-video consent. If you want stronger assurance, test the skill on a short, non-sensitive video and confirm where files are written and what gets cached before using it at scale.

Like a lobster shell, security has layers — review code before you run it.

latestvk979gwr4t2jfhtsezp55j2et7d81vszj

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

📺 Clawdis
OSLinux · macOS · Windows
Binsyt-dlp

Install

Install yt-dlp (Homebrew)
Bins: yt-dlp
brew install yt-dlp

Comments